Comparison of voiding cystourethrography and urosonography with second-generation contrast agents in simultaneous prospective study.

BACKGROUND: The invasiveness and exposure to radiation in voiding cystourethrography led to the introduction of alternative methods of diagnosis of vesicoureteral reflux, including contrast enhanced voiding urosonography. While there is a limited number of studies comparing these methods using new generation ultrasound contrast agents, none of them compared both methods simultaneously. This study is aimed at assessing agreement between contrast enhanced voiding urosonography with second-generation ultrasound contrast agents and voiding cystourethrography.
METHODS: From April 2013 to May 2014, 83 children (37 female and 46 male), mean age 3.5 years, age range from 1 month to 17.5 years, underwent prospective simultaneous assessment by contrast enhanced voiding urosonography and voiding cystourethrography, with a total of 166 uretero-renal units evaluated.
RESULTS: The sensitivity of voiding cystourethrography and contrast enhanced voiding urosonography were comparable, amounting to 88%, however, neither reached 100% for the entire studied population. The negative predictive value of voiding urosonography and voiding cystourethrography was 97%, and there was no difference between both methods.
CONCLUSION: Voiding cystourethrography and contrast enhanced voiding urosonography are comparable methods in diagnosis of vesicoureteral reflux, and can be performed alternatively. However, some limitations of contrast enhanced voiding urosonography must be remembered.

Background

Vesicoureteral reflux (VUR) is one of the most important contributors to kidney scarring, chronic kidney disease and end-stage renal disease, with the prevalence of 30–50% in pediatric and 20% in adult population(. Over the years, the number of indications for voiding cystourethrography (VCUG) has decreased as a result of its potential radiation-related side effects(. With the introduction of low-dose pulsed fluoroscopy ionizing radiation was reduced significantly, however it is still present(.

In the late 1990s, a promising alternative method for VCUG, lacking ionizing radiation, was introduced, namely contrast enhanced voiding urosonography (ce-VUS)(. While the majority of previous studies used first-generation ultrasound contrast agents (UCAs), currently only second-generation UCAs are available for clinical use. The two generations of UCAs significantly differ regarding their physical properties influencing the examination technique and results.

There is a limited number of publications that support the potential role of second-generation contrast agents in diagnosis and management of VUR in children(. Previous studies have proved that ce-VUS is a highly sensitive and radiation-free diagnostic tool for VUR imaging in children. Nonetheless the major limitation of all previous studies was a lack of simultaneous performance of both procedures, influencing the final results. In contrast, our prospective study is the first to compare the sensitivity and feasibility of both methods, VCUG and ce-VUS, in addition to the simultaneous use of second-generation UCAs.

To assess the sensitivity of ce-VUS with second generation UCAs in the diagnosis of VUR in children.

To evaluate agreement between VCUG and ce-VUS in diagnosing VUR in children.

Methods

From April 2013 to May 2014, 83 children (37 female and 46 male), mean age 3.5 years, age range from 1 month to 17.5 years underwent prospective simultaneous assessment with ce-VUS and VCUG. The examinations were performed without sedation. The study protocol was approved by the ethics committee. Parents/legal guardians were informed about the study protocol and methodology prior to the examination. Informed written consent was obtained from all patients.

The inclusion criteria for reflux examination included recurrent urinary tract infections, sonographically diagnosed dilatation of the urinary collecting system, suspicion of reflux nephropathy, follow-up of VUR and neurogenic bladder. All children were admitted by a pediatric nephrologist or urologist. All children received prophylactic antibiotic therapy in accordance to the nephrologist’s recommendations.

The patients were catheterized transurethrally under aseptic conditions with 6F–10F feeding tubes lubricated with lidocaine hydrochloride anesthetic gel (Xylocaine). After catheterization the bladder was completely emptied.

A plastic bottle containing 250 ml of saline solution (sodium chloride 0.9%) pre-warmed to 32°C, 0.5 ml of Sono Vue® and 30 ml of Visipaque (Iodixanolum, GE Healthcare A.S, Ireland) was connected through a drip system to the catheter, and placed about 80 cm above the examination table.

The solution of contrast agents for clinical use was prepared shortly before the exam under aseptic conditions. A pre-warmed plastic bottle containing 0.9% saline was filled with iodine contrast agent, followed by 0.5 ml of Sono-Vue®. The mixture was gently shaken several times, until it appeared homogenous.

Before the examination, the solution of saline and both contrast media was examined ‘in vitro’ to check for the presence of any unwanted particle formation. Immediately after reconstitution, 5 ml of solution (prepared in the same manner as for the clinical examination) were added to five Petri dishes. From each Petri dish one test sample was taken after 1min, 2 min, 3 min, 5 min, and 10 min, and viewed under microscope with standard magnification, with no precipitates seen.

The total volume of the bladder was calculated using Koff’s formula: volume in milliliters = (age in years + 2) × 30(. Filling of the bladder was performed until the child had to void, or the calculated volume was reached, or the dripping speed of infusion slowed down due to back pressure.

Ce-VUS and VCUG were performed at the same time in one cycle. The child was in the supine position, with both kidneys, the bladder, and the lower ureter scanned alternately during continued filling and voiding, using the contrast enhanced ultrasound option. The catheter was removed during voiding. VCUG was performed without continuous real-time fluoroscopic monitoring, so the person conducting ce-VUS was blind to VCUG results. Spot films were taken after filling the bladder (visible on ultrasound), and during the voiding phase.

The urethra was assessed in most cases using x-ray in an attempt to avoid additional catheterization for the ce-VUS examination, thus simultaneous assessment was technically impossible. The staff performing the procedure was protected against radiation by Personal Protective Equipment. The diagnosis of reflux was based on the appearance of high echogenicity microbubbles in the ureter or pelvicalyceal system. VCUG reflux grading was based on International System of Radiographic Grading of VUR(. In ce-VUS reflux grading system was adopted according to the system for VCUG(.

All subjects underwent a baseline gray-scale US examination of the urinary tract in the supine position. All US examinations were performed using Voluson E8 (GE Medical Systems, Milwaukee, WI, USA); 4–8 MHz convex and 7–12 linear probes or Aloka α6 (Aloka, Hitachi-Aloka Medical, Ltd, Japan); 5.0–8.0 MHz convex and 8.0–12.0 MHz linear probes.

Statistics

Since there is no “gold standard” in the diagnosis of VUR (false negatives results could occur), final diagnosis (FD) was established as the reference method. In particular, the patient was considered as having vesicoureteral reflux when the presence of VUR was detected by either of the compared methods (ce-VUS or VCUG). This approach seems to be correct, since both methods do not provide false positives results (type I error). Therefore, the analysis was reduced to calculations of sensitivity, as specificity would result as 100%.

Statistical analysis was performed with Statistica software (version 10.0; Statsoft, Inc.; Tulsa, OK, USA). The level of significance was set at α=0.05. The results of recognition by VCUG, ce-VUS and FD were presented by summary frequency tables. Agreement between VCUG and ce-VUS methods was determined by the value of Cohen’s kappa coefficient. Test of independence between the side of the kidney and the final diagnosis of VUR was implemented with the use of Yates Chi-square statistics. Sensitivity was set for VCUG and ce-VUS separately for left and right kidney, and enhanced with negative predictive values.

Results

No complications and side-effects of simultaneous intravesical application of the two contrast agents were observed.

The diagnosis of VUR was established in 33/166 uretero-renal units (20%), in 16 cases to the right kidney and in 17 to the left kidney (Tab. 1). There was no significant difference between the sides of renal reflux (p>0.05) (Tab. 2). The grading of the diagnosed refluxes is presented in Table 3.

Tab. 1

Frequency distribution of the final imaging diagnosis

	Right kidney – 0	Left kidney – 1	Total
Right kidney – 0	57	9	66
	68.7%	10.8%	79.5%
Left kidney – 1	10	7	17
	12.0%	8.4%	20.5%
Total	67	16	83
	80.7%	19.3%	100.0%

0 – vesicoureteral reflux (VUR) negative, 1 – VUR positive

Tab. 2

Independence test for the left kidney vs right kidney based on the final diagnosis of vesicoureteral reflux

	χ2 Yatesa	df	p
Male	0.29	df = 1	p = 0.592
Female	6.55	df = 1	p = 0.010

df – degrees of freedom

Tab. 3

Grading of vesicoureteral reflux diagnosed

	Grade 1	Grade 2	Grade 3	Grade 4	Grade 5
Number of URU	4	13	7	9	0

URU – uretero - renal units

VUR was not detected by ce-VUS, but identified by VCUG in 4 uretero-renal units, (3 of them were grade 2 and 1 grade 1). 4 uretero-renal units were diagnosed only in ce-VUS (2 of them were grade 2 and 2 grade 1) (Tab. 4). The distribution of the results of each method compared with the final diagnosis (FD) is described in Table 5 and Table 6.

Tab. 4

Frequency distribution of the final diagnosis of both methods: voiding cystourethrography (VCUG) and ce-VUS

	ce-VUS – 0	ce-VUS – 1	Total
VCUG – 0	133	4	137
	80.1%	2.4%	82.5%
VCUG – 1	4	25	29
	2.4%	15,06%	17.5%
Total	137	29	166
	82.5%	17.5%	100.0%

0 – vesicoureteral reflux (VUR) negative; 1 – VUR positive

Tab. 5

Summary frequency distribution of voiding cystourethrography (VCUG) results vs. final diagnosis

	Final diagnosis – 0	Final diagnosis – 1	Total
VCUG – 0	133	4	137
	80.1%	2.4%	82.5%
VCUG – 1	0	29	29
	0.0%	17.5%	17.5%
Total	133	33	166
	80.1%	19.9%	100.0%

0 – vesicoureteral reflux (VUR) negative; 1 – VUR positive

Tab. 6

Summary frequency distribution of contrast-enhanced voiding urosonography (ce-VUS) results vs. final diagnosis

	Final diagnosis – 0	Final diagnosis – 1	Total
ce-VUS – 0	133	4	137
	80.1%	2.4%	82.5%
ce-VUS – 1	0	29	29
	0.0%	17.5%	17.5%
Total	133	33	166
	80.1%	19.9%	100.0%

0 – vesicoureteral reflux (VUR) negative; 1 – VUR positive

The agreement between ce-VUS and VCUG in the diagnosis or exclusion of VUR reached more than 95% (Tab. 7), and using Landis and Koch interpretation of Cohen’s kappa coefficient (0.83) it is “nearly perfect” (.

Tab. 7

Agreement of voiding cystourethrography (VCUG) and contrast-enhanced voiding urosonography (ce-VUS) in establishing final diagnosis

Recognition	Cohen’s κ	Std. Err.	Agreement	Expected agreement	p-value
VCUG vs ce-VUS	0.833	0.078	95.18%	71.18%	0.000
VCUG vs ce-VUS for left kidney	0.881	0.109	96.39%	69.56%	0.000
VCUG vs ce-VUS for right kidney	0.779	0.109	93.98%	72.70%	0.000

Assuming that VUR detected by both methods was a true-positive result, and no reflux found by either method was representative of a true negative, the sensitivity of ce-VUS and VCUG for detection of VUR amounted to 87.9% for each method.

The negative predictive value of ce-VUS and VCUG was 97%, and there was no difference between both methods. The detailed results are presented in Table 8.

Tab. 8

Distribution of sensitivity and negative predictive value of voiding cystourethrography (VCUG) and contrast-enhanced voiding urosonography (ce-VUS)

		LK + RK	LK	RK
VCUG	Sensitivity	87.9%	100.0%	75.0%
	NPV	97.1%	100.0%	94.4%
ce-VUS	Sensitivity	87.9%	82.3%	93.7%
	NPV	97.1%	95.6%	98.5%

RK – right kidney, LK – left kidney, NPV – negative predictive value

In two cases with negative VCUG results diagnosis with ce-VUS was impossible due to the damaging of microbubbles of the contrast agent in constantly crying children. There were technical difficulties to correctly perform VCUG and ce-VUS in 5 children, mainly due to problems with achieving the voiding phase. In one case, voiding could be achieved only in the upright position, while in 4 cases spot films were not technically excellent, and not fully diagnostic during ce-VUS/VCUG.

Discussion

Ce-VUS, introduced in the early 1990s was originally thought to be a promising alternative method for classical VCUG. The short half-life and quite fast contrast destruction substantially limited the use of this method. However, the methodology of ce-VUS has changed over the last decade, especially due to the contrast agents now used for examination. Indeed, the second-generation UCAs have been recently introduced for common clinical use, replacing the older contrast agents including Levovist (Bayer-Schering Pharma, Berlin, Germany) which was commonly used in ce-VUS, but has been recently withdrawn from the market by the producer. In comparison to the second generation UCAs, the first UCA generation was not durable, and the results were highly dependent on the agent’s concentration level in the bladder(.

There are two major limitations of the use of second-generation UCAs in pediatric population, namely off-label use in the population, and the limited number of studies evaluating the sensitivity and specificity of this method(. According to the European Society of Pediatric Radiology (ESPR), standard VCUG should be performed under fluoroscopic guidance enhanced with several spot films(. The examination results in effective radiation dosage of around 0.5–0.8 mSv. In our study, we abandoned fluoroscopy, as we had continuous observation of the urinary tract in ce-VUS.

Our study found no significant difference in sensitivity between ce-VUS and VCUG, suggesting that both methods can be alternatively used with some reservations (Fig. 1, Fig. 2).

Fig. 1

A. Picture demonstrating the collecting system of the right kidney before the administration of the contrast solution into the urinary bladder. B. Reflux of the contrast agent to both collecting systems of duplicated right kidney in ce-US exam (white arrows). C. The same patient during VCUG examination (black arrows)

Fig. 2

A. Picture demonstrating the collecting system of the right kidney before the administration of the contrast solution into the urinary bladder. B. High grade reflux to the collecting systems of the right kidney in ce-US exam (white arrows). C. The same patient during VCUG examination (black arrows)

In some cases, we noticed a quick destruction of UCA due to increased intravesical pressure in children who did not cooperate and were crying. Some children had problems with postponed voiding phase, which excluded the possibility of ce-VUS use in these cases, while VCUG could still be performed. These specific UCA characteristics minimize the usefulness of this method for evaluation of the urethra during voiding, predominantly in children who are unable to cooperate.

Conclusions

VCUG and ce-VUS are comparable methods in diagnosis of VUR, and can be performed alternatively in selected groups of patients. However, some limitations of ce-VUS must be remembered, especially technical problems.

Limitations

The lack of the comparison of both methods including X-ray and ce-VUS in urethra evaluation could be viewed as a limitation of our study. We decided not to repeat the catheterization in our study to evaluate the urethra for each of the studied methods separately.